Irreversible Swelling Behavior and Reversible Hysteresis in ... - MDPI

2 downloads 0 Views 2MB Size Report
May 21, 2018 - in a DMSO/acetone mixed solvent, starting from a collapsed state in .... 1-Propanol. Ethanol. 2-Propanol. Acetone. Dioxane. Formamide.
gels Article

Irreversible Swelling Behavior and Reversible Hysteresis in Chemically Crosslinked Poly(vinyl alcohol) Gels Keiichiro Kamemaru 1 , Shintaro Usui 2 , Yumiko Hirashima 1 and Atsushi Suzuki 2, * 1 2

*

Graduate School of Education, Yokohama National University, 79-2 Tokiwadai, Hodogaya-ku, Yokohama 240-8501, Japan; [email protected] (K.K.); [email protected] (Y.H.) Graduate School of Environment and Information Sciences, Yokohama National University, 79-7 Tokiwadai, Hodogaya-ku, Yokohama 240-8501, Japan; [email protected] Correspondence: [email protected]; Tel./Fax: +81-45-339-3846

Received: 14 April 2018; Accepted: 14 May 2018; Published: 21 May 2018

 

Abstract: We report the swelling properties of chemically crosslinked poly(vinyl alcohol) (PVA) gels with high degrees of polymerization and hydrolysis. Physical crosslinking by microcrystallites was introduced in this chemical PVA gel by a simple dehydration process. The equilibrium swelling ratio was measured in several mixed solvents, which comprised two-components: a good solvent (water or dimethyl sulfoxide (DMSO)), and a poor organic solvent for PVA. In the case of aqueous/organic solvent mixtures subjected to a multiple-sample test, the swelling ratio decreased continuously when the concentration of the organic solvent increased, reaching a collapsed state in the respective pure organic solvents. In the case of DMSO, starting from a swollen state, the swelling ratio rapidly decreased by between 15 and 50 mol % when the concentration of the organic compound increased in a single-sample test. To understand the hysteresis phenomenon, the swelling ratio was measured in a DMSO/acetone mixed solvent, starting from a collapsed state in acetone. The reversibility of swelling in response to successive concentration cycles between DMSO and acetone was examined. As a result, an irreversible swelling behavior was observed in the first cycle, and the swelling ratio in acetone after the first cycle became larger than the initial ratio. Subsequently, the swelling ratio changed reversibly, with a large hysteresis near a specific molar ratio of DMSO/acetone of 60/40. The microstructures were confirmed by Fourier transform infrared spectroscopy during the cycles. The irreversible swelling behavior and hysteresis are discussed in terms of the destruction and re-formation of additional physical crosslinking in the chemical PVA gels. Keywords: poly(vinyl alcohol); chemical gel; microcrystallite; hydrogen bond; swelling behavior; hysteresis

1. Introduction Hydrogels are defined as three-dimensional, crosslinked polymer networks swollen in water, for which an infinitesimal change in environment can bring on a large volume change [1]. The backbone networks are usually formed through the chemical or physical crosslinking of polymers. In chemical hydrogels, polymer chains are connected by covalent bonds, whereas in physical hydrogels, the bonds are non-covalent. Among synthetic polymers, both chemical and physical hydrogels of poly(vinyl alcohol) (PVA) have been extensively studied for practical applications, owing to their low toxicity and high biocompatibility [2–4]. Physical PVA gels can be prepared easily by a freeze-thaw method [5–7] in water, or by a cast-drying method [8,9], starting from an aqueous PVA solution. Swelling and mechanical properties depend on the microcrystallites, which can be examined using X-ray diffraction (XRD) and Fourier Gels 2018, 4, 45; doi:10.3390/gels4020045

www.mdpi.com/journal/gels

Gels 2018, 4, 45

2 of 9

transform infrared (FT-IR) spectroscopy [5,7,10]. On the other hand, chemical PVA gels can be obtained by crosslinking PVA in solution by irradiation with γ-rays or electron beams [11,12], or by chemical reaction using crosslinking agents [13,14]. The swelling behavior of chemical PVA gels slightly crosslinked by glutaraldehyde (GA) [14–16] in mixed solvents was reported. It was found that the microcrystallites could be formed in a chemical PVA gel with high degrees of polymerization and hydrolysis during a dehydration process, which was confirmed by XRD measurements and FT-IR spectroscopy [16,17]. The formation of microcrystallites was suppressed by increasing the degree of chemical crosslinking. The swelling ratio in pure water at room temperature depends strongly on the degree of chemical crosslinking, resulting from the destruction of microcrystallites. In the case of a lower degree of chemical crosslinking density, the gel swells to a size larger than the initial volume at gelation. Moreover, the swelling ratios of the chemical PVA gels with different degrees of polymerization and hydrolysis have been measured in mixed solutions of water and organic solvents. Chemical PVA gels with high degrees of polymerization and hydrolysis assume a swollen state both in pure water and pure dimethyl sulfoxide (DMSO) (good solvents for PVA) [18], whereas these gels collapse in acetone, methanol, or ethanol (poor solvents). In addition to these simple swelling-shrinking changes, the cosolvent and cononsolvent behaviors are observed, depending on the degree of hydrolysis, temperature, and the combination of solvents [15]. In these experiments of solvent dependence, mixed solvents were prepared at different concentrations, and the samples were immersed in the respective solvents using a multiple-sample test. In order to discuss swelling behavior however, solvent concentration should be changed using one sample (single-sample test), which has not been extensively reported so far. In this study, PVA gels with high degrees of polymerization and hydrolysis were prepared by chemically crosslinking PVA using glutaraldehyde. Physical crosslinking by microcrystallites was introduced into this chemical PVA gel by a simple dehydration process. The swelling behavior of this chemical gel with physical crosslinks was examined in a two-component solvent: one is a good solvent (water or DMSO), while the other is poor (organic compound) for PVA. When the dehydrated sample was immersed in a solvent, it absorbed the solvent, swelled, and remained in its equilibrium gel state, which resulted in the destruction of weak microcrystallites. Firstly, the swelling ratio in aqueous/organic solvent mixtures with different ratios was measured by the multiple-sample test. Secondly, in order to understand the reversibility of the swelling ratio by changing the solvent composition, the swelling behavior was examined in mixed solvents of DMSO and organic solvents by the single-sample test, starting from an equilibrium state in DMSO. Finally, to explore the possibility of realizing a volume phase transition, the swelling ratio was measured in a DMSO/acetone mixed solvent using a single-sample test as a typical example of a combination of good and poor solvents. The swelling behavior is discussed in terms of the network microstructure of the gel, which was examined here by FT-IR measurements. 2. Results and Discussion 2.1. Swelling Behavior in Aqueous/Organic Mixtures The swelling ratio, d/d0 , of the chemical PVA gel in water was measured in several aqueous/organic solvent mixtures by a multiple-sample test. The gels were dehydrated after gelation and placed in their respective solvents. As shown in Figure 1, at lower concentrations of the organic solvents (alcohols, acetone, dioxane, formamide, ethanol, or DMSO) d/d0 was dependent on the solvent. At higher concentrations, d/d0 decreased continuously with an increase in concentration, reaching the collapsed state at 0.4 in the respective pure organic solvent (100 mol %), except in the case of DMSO. In the cases of propanols, d/d0 first increased and began to decrease from 10 mol %, before finally reaching a collapsed state. This exceptional change can be attributed to structural changes of bound water to the polymer [19]. The swelling behavior can be understood in terms of not only

Gels 2018, 4, 45

3 of 9

Gels 2018, 4, x FOR PEER REVIEW

3 of 9

the only solubility between the polymers and organic solvents, but also the formation and destruction of the4,solubility between Gels 2018, x FOR PEER REVIEWthe polymers and organic solvents, but also the formation and destruction 3 of 9 physical crosslinkings in the of physical crosslinkings ingels. the gels. only the solubility between the polymers and organic solvents, but also the formation and destruction of physical crosslinkings 1.6 in the gels. 1.4 1.6

1-Propanol

d / d 0 d(-)/ d 0 (-)

1.2 1.4

2-Propanol 1-Propanol DMSO

1.0 1.2

2-Propanol Ethanol

DMSO

0.8 1.0 Ethanol

0.6 0.8 0.4 0.6 0.2 0.4

Formamide

Dioxane

Acetone

0 Formamide20 Dioxane 40

Acetone

60

80

100

Concentration of Organic Solvent (mol%)

0.2

0 20 40 60 80 100 Figure 1. Swelling ratio of chemically crosslinked and dehydrated gels with different ratios of of Figure 1. Swelling ratio of chemically crosslinked and Solvent dehydrated gels with different ratios Concentration of Organic (mol%) aqueous/organic solvent mixtures at room temperature (25 °C) measured by a multiple-sample test. ◦ aqueous/organic solvent mixtures at room temperature (25 C) measured by a multiple-sample test. Figure 1. Swelling ratio of chemically crosslinked and dehydrated gels with different ratios of It should be noted that the gels showed a collapsed of d/d0 = 0.4 in aqueous/organic solvent mixtures at room temperature (25 °C) swelling measured ratio by a multiple-sample test.all the

It should be noted the gels showed a collapsed swelling of d/d0 = 0.4 solvent in all the aqueous/organic solventthat mixtures studied here, except for DMSO, i.e., forratio all two-component aqueous/organic solvent mixtures studied here, except for DMSO, i.e., for all two-component It should be noted thatsolvent the gels showed collapsed swelling ratio of d/d0 = 0.4 all solvent the mixtures containing a good (water) and a poor solvent (organic). Considering thatin DMSO is aqueous/organic solvent mixtures studied here, except for DMSO, i.e., allConsidering two-component mixtures a good solvent (water) and a poor solvent (organic). a goodcontaining organic solvent for PVA, and that d/d 0 (=1.6) is much larger thanfor that in pure water that (d/dsolvent 0 DMSO = 1.0), is mixtures containing afor good solvent (water) a poor solvent (organic). that DMSO it would be solvent interesting to measure d/dd/d 0 in 0and a(=1.6) mixed containing aConsidering poor organic solvent a good organic PVA, and that issolvent much larger than that in pure water (d/d0and =is1.0), a good organic solvent for PVA, and that d/d 0 (=1.6) is much larger than that in pure water (d/d 0 = 1.0), DMSO. Starting from an equilibrium state in DMSO (organic solvent = 0 mol %), in which it would be interesting to measure d/d0 in a mixed solvent containing a poor organic solventaand it would be sample interesting to chemical measurestate d/d0 in in a was mixed solvent a poor solvent and dehydrated of equilibrium the PVA gelDMSO placed, d/dsolvent 0containing was measured by organic ainsingle-sample test. DMSO. Starting from an (organic = 0 mol %), which a dehydrated DMSO. Starting from an equilibrium in DMSO (organic solvent = 0 mol %), in which a As is shown in Figure 2, d/d 0 initially state decreased gradually while increasing the organic solvent sample of the chemical PVA gel was placed, d/d0 was measured by a single-sample test. As is shown dehydrated sample of the chemical PVA gel was placed, d/d 0 was measured by a single-sample test. concentration, and it rapidly decreased at between and 50 mol %.organic In ordersolvent to examine the detailedand in Figure 2, d/d0 initially decreased gradually while30increasing the concentration, As is shown Figure 2, d/dby 0 initially decreased while increasing the organic solvent behavior, d/d0 in was measured a single-sample testgradually using DMSO/acetone as a typical example of the it rapidly decreased at between 30 and 50 mol %. In order to examine the detailed behavior, d/d0 was concentration, and it rapidly decreased at between 30 and mol %. In order toinexamine detailed two-component system of organic solvent and DMSO; this50has been reported the nextthe section. measured by a single-sample test using DMSO/acetone as a typical example of the two-component behavior, d/d0 was measured by a single-sample test using DMSO/acetone as a typical example of the system of organic solvent and DMSO; this has been reported in the next section. two-component system of1.6 organic solvent and DMSO; this has been reported in the next section. 1.6 1.4

d / d 0 d(-)/ d 0 (-)

Methanol

1.4 1.2

Dioxane Methanol

1.2 1.0

Dioxane 2-Propanol

1.0 0.8

Propylene Carbonate 2-Propanol

0.8 0.6 Propylene Carbonate 0.6 0.4 0 0.4

20

40

60

80

100

Concentration (mol%)

0 20 40 60 80 100 Figure 2. Swelling ratio of chemically crosslinked and dehydrated gels in DMSO-organic solvent Concentration (mol%) mixtures with increasing concentration from single-sample tests. Figure 2. Swelling ratio of chemically crosslinked and dehydrated gels in DMSO-organic solvent Figure 2. Swelling ratio of chemically crosslinked and dehydrated gels in DMSO-organic solvent mixtures with increasing concentration from single-sample tests. mixtures with increasing concentration from single-sample tests.

Gels 2018, 4, x FOR PEER REVIEW

4 of 9

Gels 2018, 4, 45

4 of 9

2.2. Swelling Ratio in DMSO/Acetone Mixed Solvent 2.2. Swelling Ratio swelling in DMSO/Acetone Mixed The equilibrium ratio, d/d 0 in aSolvent DMSO/acetone mixed solvent, was measured by varying successively acetoneswelling concentration, initially corresponded to the collapsed state of acetone The the equilibrium ratio, d/dwhich 0 in a DMSO/acetone mixed solvent, was measured by varying (d/d0 successively = 0.4). Figure showsconcentration, the equilibrium d/dinitially 0 as a function of thetoacetone concentration at 25 °C. the3acetone which corresponded the collapsed state of acetone ◦ C. During the first decreasing process, where the acetone concentration was decreased, the (d/d0 = 0.4). Figure 3 shows the equilibrium d/d0 as a function of the acetone concentration at 25 system During decreasing the acetone concentration decreased, the system remained inthe thefirst collapsed stateprocess, above where 70%, then it started to increasewas gradually, before becoming remained in the collapsed state above 70%, then it started to increase gradually, before becoming more rapid between 30 and 20 mol %, reaching the swollen state in pure DMSO. During the first more rapid between 30 and mol %,concentration reaching the swollen state in pure DMSO. first increasing process, where the 20 acetone was increased again, d/d0 During initiallythe decreased increasing process, where the acetone concentration was increased again, d/d initially decreased gradually, and then rapidly between 40 and 50 mol %, and the gel returned to0 the collapsed state in and then rapidly between 40 and 50 mol %, and the gel returned to the collapsed state pure gradually, acetone. As seen in this figure, an irreversible swelling behavior was observed, and the swelling in pure acetone. As seen in this figure, an irreversible swelling behavior was observed, and the ratio in acetone after cycling (d/d0 = 0.5) became larger than the initial ratio. Hereafter, the acetone swelling ratio in acetone after cycling (d/d0 = 0.5) became larger than the initial ratio. Hereafter, the concentration was again decreased (the second decreasing process) to that of pure DMSO, and acetone concentration was again decreased (the second decreasing process) to that of pure DMSO, increased continuously (the second increasing process), followed by the third decreasing and increased continuously (the second increasing process), followed by the third decreasingprocess. process.The swelling ratio changed continuously and reversibly, with with a large hysteresis at near thethe specific molar The swelling ratio changed continuously and reversibly, a large hysteresis at near specific ratio molar of DMSO/acetone = 60/40. ratio of DMSO/acetone = 60/40. The The diameter change of the gels was hysteresiswas wasobserved observednear near a diameter change of the gels wascontinuous, continuous,and and aa large large hysteresis a specific molar ratio. During the decreasing and increasing processes, the gels displayed reversible specific molar ratio. During the decreasing and increasing processes, the gels displayed reversible swelling behavior, except the firstdecreasing decreasing process. process. swelling behavior, except forfor the first

5

1.6

6

d / d 0 (-)

1.4 7 1st

1.2 1st

2nd

1.0 3rd

0.8

2nd

8

4

9

0.6

10 11

3

0.4

1

2

0

20

40

60

80

100

Acetone Concentration (mol%) Figure 3. Swelling ratio of of chemically and dehydrated dehydratedgels gels a DMSO/acetone mixed Figure 3. Swelling ratio chemicallycrosslinked crosslinked and in in a DMSO/acetone mixed solvent obtained a single-sample test.The Themeasurement measurement was thethe collapsed statestate in in solvent obtained by abysingle-sample test. wasinitiated initiatedfrom from collapsed acetone, concentration was successively changed changed relative and acetone. acetone, and and the the concentration was successively relativetotoDMSO DMSO and acetone.

2.3. Change in Microcrystallites DMSO/Acetone Mixed Mixed Solvent 2.3. Change in Microcrystallites in in thetheDMSO/Acetone Solvent In order to discuss the effects of destruction and re-formation of microcrystallites on swelling In order to discuss the effects of destruction and re-formation of microcrystallites on swelling behavior, ATR FT-IR spectra of the thin plate gel in the respective solvents were obtained during the behavior, ATR FT-IR spectra of the thin plate gel in the respective solvents were obtained during the first decreasing and successive increasing processes. The experimental conditions used for this thin first decreasing and successive increasing processes. The experimental conditions used for this thin plate gel were the same as those for the thin cylindrical gels used in the above diameter measurements. plate gel were the same as those for the thin cylindrical gels used in the above diameter measurements. Figure 4a shows the IR spectrum of gels with solvent during the first decreasing and successive increasing processes, which correspond to the processes shown in Figure 3. The spectra were normalized based on the C–H bending vibration at 1427 cm−1. Under these present experimental conditions, the peak area at ca. 1427 cm−1, which can be assigned to the –CH2 bending with the deformation bands of C–CH3 appearing at ca. 1377  cm−1 [20], was assumed to be equivalent in all

Gels 2018, 4, 45

5 of 9

Figure 4a shows the IR spectrum of gels with solvent during the first decreasing and successive increasing processes, which correspond to the processes shown in Figure 3. The spectra were normalized based on the C–H bending vibration at 1427 cm−1 . Under these present experimental Gels 2018, 4, x FOR PEER REVIEW 5 of 9 conditions, the peak area at ca. 1427 cm− 1 , which can be assigned to the –CH2 bending with the deformation bands of C–CH3 appearing at ca. 1377 cm− 1 [20], was assumed to be equivalent in spectra, and the spectra were normalized by the respective peak areas. The IR spectra of pure DMSO all spectra, and the spectra were normalized by the respective peak areas. The IR spectra of pure and acetone are added to this figure. The peaks located at ca. 1360 and 1210 cm–1 originate in the DMSO and acetone are added to this figure. The peaks–1located at ca. 1360 and 1210 cm−1 originate presence of acetone, while those at ca. 1050 and 900 cm in the presence of DMSO. The large peak at in the presence of−1acetone, while those at ca. 1050 and 900 cm−1 in the presence of DMSO. The large around 3300 cm consists of two components, which is assigned to O–H stretching of the nonpeak at around 3300 cm−1 consists of two components, which is assigned to O–H stretching of the hydrogen-bonded (3409 cm–1) and hydrogen-bonded O–H groups (3295 cm–1) [17]. As shown in this non-hydrogen-bonded (3409 cm−1 ) and hydrogen-bonded O–H groups (3295 cm−1 ) [17]. As shown figure, the peak shifts to higher wavenumbers in the first decreasing process (No. 1 to 5), and to lower in this figure, the peak shifts to higher wavenumbers in the first decreasing process (No. 1 to 5), wavenumbers in the first increasing process (No. 5 to 11). The peak height shows a tendency to and to lower wavenumbers in the first increasing process (No. 5 to 11). The peak height shows decrease and then increase during this cycle. These observations indicate that the ratio of the numbers aoftendency to decrease andO–H thengroups increase duringand thisdecreased, cycle. These observations indicate that the free to hydrogen bonded increased since the hydrogen bonds between ratio of the numbers of free to hydrogen bonded O–H groups increased and decreased, since the polymer chains were respectively destroyed and re-formed by changing the DMSO/acetone ratio. On hydrogen bonds the between chains were respectively destroyed anddue re-formed by changing the –1 corresponding the other hand, peak polymer at 1143 cm to C–O stretching to the microcrystallites − 1 DMSO/acetone ratio. On the other hand, the peak at 1143 cm corresponding to C–O stretching due [21–23] disappears in the swollen state and re-appears in the collapsed phase. Figure 4b shows that to microcrystallites [21–23] in the swollen state and re-appears in thethe collapsed –1 wasdisappears thethe peak area at 1143 cm obtained by deconvoluting the spectra to remove effects phase. of the − 1 was obtained by deconvoluting the spectra to remove Figure 4b shows that the peak area at 1143 cm –1 other peaks. The peak area at 1143 cm decreases in the first decreasing process and increases in the 1 decreases in the first decreasing process and the effects ofincreasing the other peaks. The resulting peak area at 1143the cm−decrement successive process, from and increment of the ratio of increases in the successive increasing process, the the decrement and and increment of theinratio microcrystallites to amorphous networks, orresulting in other from words, decrement increment the of microcrystallites to amorphous networks, or in other words, the decrement and increment in the degree of crystallization, respectively. There seems to be a discrepancy between the swelling ratio degree of crystallization, respectively. There seems to be4 aisdiscrepancy between the swelling ratio and and peak area. For example, the peak area of No. almost zero in Figure 4b, although the peak area. For example, the peak area of No. 4 is almost zero in Figure 4b, although the corresponding corresponding swelling ratio does not reach the swollen state. This discrepancy might result from swelling ratio not reach the conditions, swollen state. This result fromof deficiencies in the deficiencies indoes the experimental such as discrepancy those due tomight the evaporation solvent water at experimental conditions, such as those due to the evaporation of solvent water at the surface. the surface. Acetone

11

Absorbance (a.u.)

DMSO 10

11

10

9

9

9

8 7

8

8

6

7

7 6 5

5

6

4

5

4

3

10

4 3

3

2

2

1

1

4000

11

3500

[A]

3000

2500

2000

1500

–1

Wavenumber (cm ) (a) Figure 4. Cont.

1000

[B]

3600

3400

[A]

3200

2 1 1160

1140

[B]

1120

Gels 2018, 4, 45

6 of 9

–1

Normalized peak area at 1143cm (a.u.)

Gels 2018, 4, x FOR PEER REVIEW

6 of 9

1

0.4 11 1st

0.3

9

10

2

0.2

1st

3 8

0.1 5

0.0 0

4 6

7

20 40 60 80 Acetone Concentration (mol%) (b)

100

Figure 4. 4. (a) in various various Figure (a) ATR ATR FT/IR FT/IR spectra spectra of of aa chemically chemically crosslinked crosslinked and and dehydrated dehydrated gel gel in DMSO/acetone mixtures detailed changes at DMSO/acetone mixturesduring duringthe thesingle-sample single-sampletest testshown shownininFigure Figure3.3.The The detailed changes −1 (region [A]) and 1143 cm−1 (region [B]) are displayed; (b) The peak area at 1143 cm−1 around 3300 cm − 1 − 1 at around 3300 cm (region [A]) and 1143 cm (region [B]) are displayed; (b) The peak area at −1, obtained by deconvoluting the spectra to remove the normalized by the peakby area ca. 1427 cmca. 1143 cm− 1 normalized theat peak area at 1427 cm−1 , obtained by deconvoluting the spectra to effects of other peaks. The respective number 1 tofrom 11 corresponds to those Figure 3. The remove the effects of other peaks. The respective from number 1 to 11 corresponds to in those in Figure 3. acetone concentrations are as No. No. 1: 100, No. No. 2: 50, 3: 40, 4: 30, 5: 20, 6: 30, The acetone concentrations arefollows; as follows; 1: 100, 2: No. 50, No. 3: No. 40, No. 4: No. 30, No. 5: No. 20, No. 6: No.No. 7: 40, 8: 50, 9: 60, 10: 70, %. %. 30, 7: No. 40, No. 8: No. 50, No. 9: No. 60, No. 10:No. 70, 11: No.100 11: mol 100 mol

2.4. Origin Origin of of Reversible Reversible Change Change with with aa Large Large Hysteresis Hysteresis 2.4. Two important important observations observations can can be be made made about about the the observed observed swelling swelling behavior behavior of of chemically chemically Two crosslinked PVA gels in response to solvent concentration changes in DMSO/acetone mixed crosslinked PVA gels in response to solvent concentration changes in DMSO/acetone mixed solutions. solutions. is is that is an irreversible swellingduring behavior the first decreasing and One is thatOne there anthere irreversible swelling behavior the during first decreasing and increasing increasing processes. This irreversible swelling behavior suggests that microcrystallites are destroyed processes. This irreversible swelling behavior suggests that microcrystallites are destroyed during during first decreasing not completely the first the first the decreasing process, process, and theyand werethey not were completely re-formed re-formed during theduring first increasing increasingInprocess. In othermicrocrystallites words, microcrystallites can recover, the recovered amount is smaller process. other words, can recover, but the but recovered amount is smaller than than the initial amount. According to a report on the ionized gels of poly(N-isopropylacrylamide) the initial amount. According to a report on the ionized gels of poly(N-isopropylacrylamide) [24] [24] poly(sodium and poly(sodium acrylate) where re-swelling transitions examined, hydrogen and acrylate) [25], [25], where re-swelling transitions were were examined, hydrogen bondsbonds were were destroyed in response to the changes in the external conditions, and could not be reformed; destroyed in response to the changes in the external conditions, and could not be reformed; furthermore, furthermore, volume macroscopic volume changes weretransition. a one-way In transition. of thePVA present macroscopic changes were a one-way the caseIn ofthe thecase present gels,PVA the gels, the destruction of microcrystallites by increasing the temperature was reported earlier [16,17]. destruction of microcrystallites by increasing the temperature was reported earlier [16,17]. It was also was also transition, a one-way transition, and re-formation of microcrystallites was not detected, the aIt one-way and re-formation of microcrystallites was not detected, although although the swelling swelling ratio decreased slightly when the temperature decreased. In the present case, the gels ratio decreased slightly when the temperature decreased. In the present case, the gels returned to to atostate close to the collapsed upon the re-formation of microcrystallites, which were areturned state close the collapsed phase upon phase the re-formation of microcrystallites, which were detected for detected for the first time. This evidence suggests that there are two kinds of microcrystallites: stable the first time. This evidence suggests that there are two kinds of microcrystallites: stable and unstable, and reversible unstable, i.e., reversible or irreversible changes response change in solvent composition. i.e., or irreversible changes in response toin the change to in the solvent composition. The other important observation is that d/d 0 changed reversibly with a large hysteresis after the The other important observation is that d/d0 changed reversibly with a large hysteresis after the second decreasing decreasing process. process. The The swelling swelling behavior behavior was was reversible reversible in in the the sense sense that thatd/d d/d0 at high and second 0 at high and low acetone cycles, suggesting that thethe destruction andand relow acetone concentration concentrationwas wasthe thesame sameininsubsequent subsequent cycles, suggesting that destruction formation of microcrystallites that survived the first cycle were reversible in response to a change in re-formation of microcrystallites that survived the first cycle were reversible in response to a change in solvent concentration. The large hysteresis is attributed to the formation of microcrystallites in the solvent concentration. The large hysteresis is attributed to the formation of microcrystallites in the system; once oncethethe microcrystallites are destroyed the decreasing process, subsequent system; microcrystallites are destroyed duringduring the decreasing process, subsequent increasing increasing does not cause them to form until a much higher acetone concentration is used. does not cause them to form until a much higher acetone concentration is used. Finally,ititshould should noted the destruction and re-formation of microcrystallites was Finally, be be noted thatthat the destruction and re-formation of microcrystallites was detected detected in the present system, which resulted in the hysteresis phenomenon. Hysteresis can in the present system, which resulted in the hysteresis phenomenon. Hysteresis can be found be in found in many materials systems, such as ferromagnetic, ferroelectric, and viscoelastic materials. Comparing with those phenomena, the present hysteresis was observed in the equilibrium state,

Gels 2018, 4, 45

7 of 9

many materials systems, such as ferromagnetic, ferroelectric, and viscoelastic materials. Comparing with those phenomena, the present hysteresis was observed in the equilibrium state, which is rate-independent, similar to magnetic hysteresis loops. In hydrogel systems, on the other hand, this phenomenon is reminiscent of multiple phase transitions [26], which are characterized by discontinuous and reversible changes in swelling ratios with large hysteresis, induced by a competitive balance between hydrogen bonds and hydrophobic interactions. Considering the hypothesis of universality of the phase transition of chemical gels, it might be possible to realize a reversible and discontinuous volume change in the present system. For that purpose, it would be necessary to adjust the material or environmental parameters to induce a positive osmotic pressure, and increase the swelling pressure in the good solvent relative to the destruction force of the microcrystallites. This remains a challenging topic in this field. 3. Conclusions Chemical PVA gels were prepared by crosslinking PVA slightly with glutaraldehyde, washed with a large amount of pure water, and dehydrated fully in air at room temperature to introduce physical crosslinking. Both irreversible and reversible swelling behavior and hysteresis were observed in these PVA gels. The gel reached its collapsed state in a pure organic solvent by a multiple-sample test, but did not reach the same state in a single-sample test. The equilibrium swelling ratios in a DMSO/acetone mixed solution were measured at room temperature to examine the relation between macroscopic swelling ratio and microscopic network structure. The swollen gel in DMSO could shrink in acetone, but did not return to the initial collapsed state after the gel had experienced the swollen state. This irreversible swelling behavior was attributed to the destruction of the microcrystallites formed during the initial dehydration after gelation. However, the swelling ratio changed continuously and reversibly, with a large hysteresis at around a specific molar ratio of DMSO/acetone of 60/40. This was consistent with the change in the microcrystallites observed by FT-IR measurements; microcrystallites were destroyed in DMSO and re-formed in acetone. This irreversible swelling behavior could not be observed in physical gels, although the sizes and amounts of microcrystallites were similar to those of the chemical gels. These observations were attributed to the irreversible destruction of weak microcrystallites, and to the reversible destruction and re-formation change of the strong microcrystallites. Finally, it was concluded that the destruction and re-formation of microcrystallites in chemical PVA gels were reversibly controlled by concentration changes in the solvent, which was assisted by the chemical crosslinking. It is important to note that the present findings can be confirmed by other measurements, using XRD, DSC, and other techniques, which merit further research. 4. Experimental 4.1. Sample Preparation Chemically crosslinked PVA gels with high degrees of polymerization and hydrolysis were obtained by crosslinking PVA solutions with GA. The PVA powder was supplied by Kuraray Co., Ltd. (Cat. No. PVA117, Tokyo, Japan), and was used without further purification. The average degree of polymerization was 1700, and the average degree of hydrolysis was between 98 and 99 mol %. The PVA powder was dissolved in deionized and distilled water at 90 ◦ C or higher for more than 2 h. Gels were prepared in two shapes: thin cylindrical gels for the measurement of swelling ratio were prepared in glass microcapillaries, with an inner diameter of 1.44 mm and inner volume of 50 µL, and thin plate gels for FT-IR measurements were prepared between two slides with a thin spacer (thickness: about 2 mm). The concentration of the monomer units of PVA was fixed at ca. 2100 mM, and the proportion of GA (Wako Pure Chemical Industries, Ltd. (Tokyo, Japan), 25% aqueous solution) to the total concentration of monomer units of PVA plus the monomer of GA was 0.5 mol %. It is noteworthy

Gels 2018, 4, 45

8 of 9

that the degree of crosslinking of 0.5 mol % GA corresponds to a chain length of about 100 monomer units of PVA. After mixing GA in dissolved PVA solution, 0.3 mL of 1 M HCl was added to the mixture as a catalyst. The solution was placed in a furnace at 30 ◦ C for 48 h. After gelation, the cylindrical and plate gels were removed from the microcapillaries and mold, and subsequently washed in pure water (the water volume was 1000 times larger than the gel volume) for 48 h, to wash away any residual chemicals or uncrosslinked polymers from the polymer networks. Then, the gels were again placed in an oven at 30 ◦ C for 48 h to dry. 4.2. Measurements of Swelling Ratio and FT-IR The equilibrium diameter of a cylindrical gel, d, was measured by phase contrast microscopy. A sample was placed in respective solvents and left for more than 48 h, typically 96 h, which was sufficient for the gel to reach its equilibrium state. The swelling degree of the gel is represented by the swelling ratio, d/d0 , where d0 is the gelation diameter (the inner diameter of the glass capillary, 1.44 mm). To confirm the structural details of the gels, FT-IR measurements using the attenuated total reflectance (ATR) technique was performed on the dehydrated plate gels at room temperature. We used an FT-IR spectrophotometer (Jasco: FT/IR-610, Tokyo, Japan) equipped with an ATR (attenuated total reflectance) attachment with a ZnSe crystal. An appropriate amount of the gel was placed on the ZnSe crystal, and the FT-IR spectra were then recorded at room temperature. Author Contributions: K.K. and A.S. conceived and designed the experiments; K.K. and S.U. performed the experiments and analyzed the data; Y.H. optimized the method for preparing chemically crosslinked poly(vinyl alcohol) gels and supported the analysis of the FT-IR spectra. Acknowledgments: We would like to express our sincere gratitude to Kuraray Co., Ltd. (Tokyo, Japan) for supplying PVA powders. This work was supported in part by Japan Society for the Promotion of Science: JSPS KAKENHI grant numbers JP23000011 and JP16H03170. Conflicts of Interest: The authors declare no conflict of interest.

References 1. 2. 3. 4. 5. 6. 7. 8. 9.

10. 11.

Tanaka, T. Gels. Sci. Am. 1981, 244, 124–136. [CrossRef] [PubMed] Oka, M.; Ushio, K.; Kumar, P.; Ikeuchi, K.; Hyon, S.H.; Nakamura, T.; Fujita, H. Development of artificial cartilage. Proc. Inst. Mech. Eng. H J. Eng. Med. 2000, 214, 59–68. [CrossRef] [PubMed] Kobayashi, M.; Chang, Y.S.; Oka, M. A two year in vivo study of polyvinyl alcohol-hydrogel (PVA-H) artificial meniscus. Biomaterials 2005, 26, 3243–3248. [CrossRef] [PubMed] Baker, M.I.; Walsh, S.P.; Schwartz, Z.; Boyan, B.D. A review of polyvinyl alcohol and its uses in cartilage and orthopedic applications. J. Biomed. Mater. Res. B 2012, 100, 1451–1457. [CrossRef] [PubMed] Hassan, C.M.; Peppas, N.A. Structure and applications of poly(vinyl alcohol) hydrogels produced by conventional crosslinking or by freezing/thawing methods. Adv. Polym. Sci. 2000, 153, 37–65. Nambu, M. Japanese Patent Kokai. Japan Patent 59/56446, 1984. Lozinsky, V.I. Cryotropic gelation of poly(vinyl alcohol). Russ. Chem. Rev. 1998, 67, 573–586. [CrossRef] Otsuka, E.; Suzuki, A. A simple method to obtain a swollen PVA gel crosslinked by hydrogen bonds. J. Appl. Polym. Sci. 2009, 114, 10–16. [CrossRef] Otsuka, E.; Suzuki, A. Swelling properties of physically cross-linked PVA gels prepared by a cast-drying method. In Gels: Structures, Properties, and Functions; Springer: Berlin/Heidelberg, Germany, 2009; pp. 121–126. Otsuka, E.; Sugiyama, M.; Suzuki, A. Network microstructure of PVA cast gels observed by SAXS measurements. J. Phys. Conf. Ser. 2010, 247, 012043. [CrossRef] Wang, B.; Mukataka, S.; Kodama, M.; Kokufuta, E. Viscometric and light scattering studies on microgel formation by γ-ray irradiation to aqueous oxygen-free solutions of poly(vinyl alcohol). Langmuir 1997, 13, 6108–6114. [CrossRef]

Gels 2018, 4, 45

12. 13. 14. 15. 16. 17. 18. 19.

20. 21. 22. 23. 24. 25. 26.

9 of 9

Yoshii, F.; Makuuchi, K.; Darwis, D.; Iriawan, T.; Razzak, M.T.; Rosiak, J.M. Heat resistance poly(vinyl alcohol) hydrogel. Radiat. Phys. Chem. 1995, 46, 169–174. [CrossRef] Praptowidodo, V.S. Influence of swelling on water transport through PVA-based membrane. J. Mol. Struct. 2005, 739, 207–212. [CrossRef] Dai, W.S.; Barbari, T.A. Hydrogel membranes with mesh size asymmetry based on the gradient crosslinking of poly(vinyl alcohol). J. Membr. Sci. 1999, 156, 67–79. [CrossRef] Kudo, S.; Otsuka, E.; Suzuki, A. Swelling behavior of chemically crosslinked PVA gels in mixed solvents. J. Polym. Sci. B 2010, 48, 1978–1986. [CrossRef] Otsuka, E.; Kudo, S.; Sugiyama, M.; Suzuki, A. Effects of microcrystallites on swelling behavior in chemically crosslinked poly(vinyl alcohol) gels. J. Polym. Sci. B 2011, 49, 96–102. [CrossRef] Otsuka, E.; Sugiyama, M.; Suzuki, A. Formation and destruction of physical crosslinks by mild treatments in chemically crosslinked poly(vinyl alcohol) gels. Polym. Bull. 2011, 67, 1215–1226. [CrossRef] Hirokawa, Y.; Tanaka, T. Volume phase transition in a non-ionic gel. J. Chem. Phys. 1984, 81, 6379–6380. [CrossRef] Shinyashiki, N.; Shimomura, M.; Ushiyama, T.; Miyagawa, T.; Yagihara, S. Dynamics of Water in Partially Crystallized Polymer/Water Mixtures Studied by Dielectric Spectroscopy. J. Phys. Chem. B 2007, 111, 10079–10087. [CrossRef] [PubMed] Nkhwa, S.; Lauriaga, K.F.; Kemal, E.; Deb, S. Poly(vinyl alcohol): Physical Approaches to Designing Biomaterials for Biomedical Applications. Conf. Pap. Sci. 2014, 2014, 403472. [CrossRef] Miya, M.; Iwamoto, R.; Mima, S. FT-IR study of intermolecular interactions in polymer blends. J. Polym. Sci. B 1984, 22, 1149–1151. [CrossRef] Mallapragada, S.K.; Peppas, N.A. Dissolution mechanism of semicrystalline poly(vinyl alcohol) in water. J. Polym. Sci. B 1996, 34, 1339–1346. [CrossRef] Lee, J.; Lee, K.J.; Jang, J. Effect of silica nanofillers on isothermal crystallization of poly(vinyl alcohol): In situ ATR-FTIR study. Polym. Test. 2008, 27, 360–367. [CrossRef] Hirashima, Y.; Suzuki, A. Roles of Hydrogen Bonding on the Volume Phase Transition of Ionized Poly(N-isopropylacrylamide) Gels. J. Phys. Soc. Jpn. 2004, 73, 404–411. [CrossRef] Sato, H.; Hirashima, Y.; Suzuki, A. Reswelling transition of poly(sodium acrylate) gels due to destruction of hydrogen bonds observed by ATR FTIR spectroscopy. J. Appl. Polym. Sci. 2007, 105, 3809–3816. [CrossRef] Annaka, M.; Tanaka, T. Multiple phases of polymer gels. Nature 1992, 355, 430–432. [CrossRef] © 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).